Two-Stage Clutter Suppression Method for Human Detection Using FMCW Radar

Human detection and localization play an important role in smart home monitoring systems. Generally, the signal from a human subject can be detected by capturing the micro-movement arising from vital signs. However, in an indoor environment, there are diverse vibrating electronic devices such as a fan, an air conditioner, and a desktop computer, which can impede the detection of the human subject. These non-static clutters have high variation and can be falsely detected as human subjects. To cope with this problem, we propose a two-stage clutter suppression method using a frequency-modulated continuous wave (FMCW) radar to suppress both static and non-static clutters. First, the micro-movement information is used to suppress static clutters. Next, using the velocity information estimated from the FMCW radar, four types of metrics (sidelobe power, peak-to-average power ratio, Gini coefficient, and entropy) are calculated which are used as the criterion to distinguish human subjects from non-static clutters. Experimental results verified that the proposed method can accurately detect the position of a human subject, while successfully suppressing both static and non-static clutters.